Classifying neurons into functional subgroups is essential to the understanding of function in the central nervous system which has a large number of heterogeneous populations of neurons. In the somatosensory system, many attempts have been made to classify neurons both in the peripheral and central nervous system. However, none of them satisfactorily represent functions of the somatosensory system. The most important function of the somatosensory system is to transmit various sensory modalities and submodalities of somatic sensation. Therefore, the most useful functional classification of the somatosensory system would be grouping neurons by their potential contribution to sensory modalities. It is hypothesized that objective statistical treatment of the responses of somatosensory neurons to a set of stimuli which reflect sensory modalities will lead us to a new and functionally relevant classification scheme for somatosensory neurons. The goal of the present proposal is to test this hypothesis and thereby establish a new objective means of classifying somatosensory neurons. To achieve this goal, the following specific aims are proposed. First aim is to develop a "standard set of stimuli" to apply to the receptive fields of all the somatosensory neurons whose responses can then be used as the basis for a new classification scheme. And then, aims are directed to perform actual classification of somatosensory neurons of the rat and monkey at 4 different levels: 1) primary afferent neurons; 2) dorsal horn neurons; 3) spinothalamic tract neurons; 4) somatosensory neurons in the thalamus. To achieve these aims, experiments will be performed in anesthetized rats and monkeys. Single cell recordings will be made from somatosensory neurons at the above 4 different levels. Multivariate statistical analysis, cluster analysis, will be performed on responses of neurons to the "standard set of stimuli" to classify neurons objectively. Achieving these specific aims will undoubtedly lead to a better understanding of sensory processing of somatosensory information in normal as well as in diseased states. This is the long term objective of the present proposal. Establishing an objective and functionally meaningful grouping of somatosensory neurons in normal animals will help us to interpret changes seen in diseased states.